Phenological responses to temperature are similar across spatial and temporal gradients - lessons learned from European long-term data.

Understanding plant phenology patterns under global warming is important because they influence many processes, from specific dynamics, such as species interactions, to ecosystem functions, for instance, carbon uptake. Studying phenological changes requires long-term data covering the species' distribution. However, such data is typically only locally available and restricted to certain phenophases, e.g., flowering. One common alternative is to use the space-for-time substitution (SFTS) approach. However, this method relies on the rarely tested space-time equivalence assumption. In this study, we assessed the space-time equivalence of flowering phenology by comparing its response to temperature along temporal and spatial gradients. Furthermore, we evaluated the influence of recent climate change on this assumption, analyzing the responses between 1950-1985 and 1986-2022. We used mean annual temperature data from the Era5-Land product and long-term flowering onset records of 23 herbaceous and dwarf shrub species from the PEP725 database. We found that flowering responses to temperature were mostly negative, and the responses were 67% similar between both gradients. However, the response variation along the spatial gradient did not represent the total variation observed in the temporal one. The similarity in responses between both gradients was higher (73.6%) in warmer decades (73.6% in 1986-2022 compared to 65.8% in 1950-1985), suggesting that high temperatures may synchronize phenology across space and time. We conclude that the SFTS predicts the phenological response direction to the temperature correctly. However, we recommend considering that the use of spatial gradients may not reveal the full expected response variation to temperature in a temporal gradient, especially when making general conclusions from species-specific inferences.